The eurozone, as designed, has failed. It was based on a set of principles that have proved unworkable at the first contact with a financial and fiscal crisis.
It has only two options: to go forwards towards a closer union or backwards towards at least partial dissolution. This is what is at stake.
The eurozone was supposed to be an updated version of the classical gold standard. Countries in external deficit receive private financing from abroad. If such financing dries up, economic activity shrinks. Unemployment then drives down wages and prices, causing an "internal devaluation".
In the long run, this should deliver financeable balances in the external payments and fiscal accounts, though only after many years of pain. In the eurozone, however, much of this borrowing flows via banks.
When the crisis comes, liquidity-starved banking sectors start to collapse. Credit-constrained governments can do little, or nothing, to prevent that from happening.
This, then, is a gold standard on financial sector steroids. The role of banks is central. Almost all of the money in a contemporary economy consists of the liabilities of financial institutions.
In the eurozone, for example, currency in circulation is just 9 percent of broad money. If this is a true currency union, a deposit in any eurozone bank must be the equivalent of a deposit in any other bank.
But what happens if the banks in a given country are on the verge of collapse? The answer is that this presumption of equal value no longer holds. A euro in a Greek bank is today no longer the same as a euro in a German bank.
In this situation, there is not only the risk of a run on a bank but also the risk of a run on a national banking system. This is, of course, what the federal government has prevented in the US.
At last month's Munich economic summit, Hans-Werner Sinn, president of the Ifo Institute for Economic Research, brilliantly elucidated the implications of the response to this threat of the European System of Central Banks (ESCB). The latter has acted as lender of last resort to troubled banks.
But, because these banks belonged to countries with external deficits, the ESCB has been indirectly financing those deficits, too. Moreover, because national central banks have lent against discounted public debt, they have been financing their governments. Let us call a spade a spade: this is central bank finance of the state.
The ESCB's finance flows via the euro system's real-time settlement system ("target-2"). Huge asset and liability positions have now emerged among the national central banks, with the Bundesbank the dominant creditor.
Indeed, Prof Sinn notes the symmetry between the current account deficits of Greece, Ireland, Portugal and Spain and the cumulative claims of the Bundesbank upon other central banks since 2008 (when the private finance of weaker economies dried up).
Government insolvencies would now also threaten the solvency of debtor country central banks. This would then impose large losses on creditor country central banks, which national taxpayers would have to make good.
This would be a fiscal transfer by the back door. Indeed, that this is likely to happen is quite clear from the striking interview with Lorenzo Bini Smaghi, a member of the board of the European Central Bank, in the FT of May 29 2011.
Prof Sinn makes three other points. First, this backdoor way of financing debtor countries cannot continue for very long.
By shifting so much of the eurozone's money creation towards indirect finance of deficit countries, the system has had to withdraw credit from commercial banks in creditor countries.
Within two years, he states, the latter will have negative credit positions with their national central banks - in other words, be owed money by them. For this reason, these operations will then have to cease.
Second, the only way to stop them, without a crisis, is for solvent governments to take over what are, in essence, fiscal operations. Yet, third, when one adds the sums owed by national central banks to the debts of national governments, totals are now frighteningly high.
The only way out is to return to a situation in which the private sector finances both the banks and the governments. But this will take many years, if it can be done with today's huge debt levels at all.
Debt restructuring looks inevitable. Yet it is also easy to see why it would be a nightmare, particularly if, as Mr Bini Smaghi insists, the ECB would refuse to lend against the debt of defaulting states.
In the absence of ECB support, banks would collapse. Governments would surely have to freeze bank accounts and redenominate debt in a new currency. A run from the public and private debts of every other fragile country would ensue. That would drive these countries towards a similar catastrophe.
The eurozone would then unravel. The alternative would be a politically explosive operation to recycle fleeing outflows via public sector inflows. Events have, in short, thoroughly falsified the premises of the original design. If that is the design the dominant members still want, they must remove some of the existing members. Managing that process is, however, nigh on impossible.
If, however, they want the eurozone to work as it is, at least three changes are inescapable. First, banking systems cannot be allowed to remain national. Banks must be backed by a common treasury or by the treasury of unimpeachably solvent member states. Second, cross-border crisis finance must be shifted from the ESCB to a sufficiently large public fund. Third, if the perils of sovereign defaults are to be avoided, as the ECB insists, finance of weak countries must be taken out of the market for years, perhaps even a decade.
Such finance must be offered on manageable conditions in terms of the cost but stiff requirements in terms of the reforms. Whether the resulting system should be called a "transfer union" is uncertain: that depends on whether borrowers pay everything back (which I doubt).
But it would surely be a "support union". The eurozone confronts a choice between two intolerable options: either default and partial dissolution or open-ended official support.
The existence of this choice proves that an enduring union will at the very least need deeper financial integration and greater fiscal support than was originally envisaged.
How will the politics of these choices now play out? I truly have no idea. I wonder whether anybody does.